Hierarchical Content Delivery Cache Grouping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Content delivery networks (CDNs) face challenges in efficiently delivering content objects due to cache misses, where edge servers lack the required content, leading to increased latency and reduced quality of service, especially when content is not available locally and requires retrieval from external origin servers.
Innovation Solution
A hierarchical content retrieval system where edge servers pass information to parent servers to locate content objects, escalating through a chain of servers until the origin server is reached, with optional caching or hosting at each server in the chain to facilitate delivery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If edge servers cache content locally to reduce latency, then delivery speed improves, but cache memory capacity requirements increase
Solution Approach 1:
The system segments the content delivery function across multiple hierarchical levels (edge servers, parent servers, origin servers). Instead of requiring each edge server to cache all possible content, the system divides content into different cache levels, where edge servers cache frequently accessed content locally while less frequently accessed content is stored at parent or origin servers. This segmentation resolves the contradiction by enabling fast local delivery for popular content while avoiding the need for massive cache capacity at every edge server.
Solution Approach 2:
The patent introduces a hierarchical dimension to content delivery, organizing servers into multiple levels (edge, parent, origin) rather than a flat architecture. This vertical dimension allows the system to manage cache capacity effectively by distributing content across levels, where each level handles different portions of the content hierarchy, thereby reducing the cache burden at individual edge servers while maintaining fast delivery.
2Reliability
If edge servers store all possible content objects, then content availability improves, but device complexity and storage requirements increase
Solution Approach 1:
The system segments content storage responsibilities across hierarchical server levels. Edge servers store only frequently accessed content locally, while parent servers and origin servers store the complete content repository. This segmentation maintains content availability by ensuring that if an edge server lacks a particular object, the request can be fulfilled by parent or origin servers, avoiding the need for every edge server to store all possible content.
Solution Approach 2:
Parent servers act as intermediaries between edge servers and origin servers. When an edge server cannot find content locally, it queries parent servers, which may have cached the content or can retrieve it from the origin server. This intermediary layer simplifies the system by allowing edge servers to maintain simple local caches while ensuring content availability through the hierarchical mediation structure.
3Measurement precision
If the system queries multiple servers to locate content, then content location accuracy improves, but search time and latency increase
Solution Approach 1:
The system performs preliminary content placement and caching at hierarchical levels before requests arrive. Parent servers and origin servers pre-load frequently accessed content into edge server caches, and the hierarchical structure itself is pre-configured with knowledge of content locations. When a request arrives, the edge server can quickly determine whether content is locally cached or needs to be retrieved from parent/origin servers, reducing search time while maintaining accurate content location through the pre-established hierarchy.
Solution Approach 2:
The hierarchical architecture adds a vertical dimension to content location, organizing servers into levels (edge, parent, origin) that are pre-ordered by proximity to the content and by frequency of access. This dimensional organization allows the system to efficiently navigate content locations by first checking local caches, then parent servers, and finally origin servers, reducing the time required to locate content compared to random or flat-search approaches.
4Adaptability or versatility
If content is retrieved from external origin servers, then content diversity improves, but delivery latency and QoS degradation increase
Solution Approach 1:
The system segments content delivery into local fulfillment (from edge server cache) and remote fulfillment (from parent or origin server). Frequently accessed content is segmented and cached at the edge server level for immediate delivery, while less frequently accessed content remains at parent or origin servers. This segmentation allows the system to maintain content diversity by retrieving from external servers when necessary while minimizing latency for the majority of requests that can be satisfied locally.
Solution Approach 2:
The system dynamically adjusts cache policies and retrieval parameters based on content popularity and access patterns. By monitoring which content objects are frequently requested, the system changes cache placement decisions and retrieval strategies to prioritize local cache hits for popular content, thereby reducing latency for those objects while maintaining the ability to retrieve diverse content from origin servers when needed.
Data Source
AI summary
Content delivery networks (CDNs) deliver content objects for others is disclosed. End user computers are directed to an edge server for delivery of a requested content object by a universal resource indicator (URI). When an edge server does not have a copy of the content object from the URI, information is successively passed to ancestor servers within a hierarchy until the content object is found. There can be different hierarchies designated for different URIs or times at which requests are received. Once the content object is located in the hierarchical chain, the content object is passed back down the chain to the edge server for delivery.